Seattle News

19-11-2026

Oregon Sues OpenAI for $2.035 Billion Over ChatGPT Risks to Children

A lawsuit filed on November 17, 2026, by Oregon Attorney General Dan Rayfield is one of the most sweeping attempts by a state to hold ChatGPT’s creator accountable for potential harm to minors. The state is seeking at least $2.035 billion and a court order requiring OpenAI to do a better job of determining users’ ages, alerting parents and restricting dangerous conversations about suicide, self-harm, eating disorders and violence. The company rejects the accusation that it failed to act, saying it is already developing parental controls, age-based safeguards and other protections.

On Tuesday, November 17, Oregon filed a lawsuit against six OpenAI entities in Multnomah County Circuit Court in Portland. Behind the dry legal language is an unusually broad claim: this is not about one specific incident in Oregon or a private dispute between a family and a technology company. Rather, the state argues that the way ChatGPT is designed and promoted poses a public threat to children. Rayfield alleges that OpenAI knew about the risks of emotional dependence and harmful responses but continued to present the product to parents as safe.

The $2.035 billion is not a fine already imposed by a court, but the minimum amount in damages the state intends to seek at trial. In its complaint, Oregon cites the costs of addressing harm to residents and also seeks compensation for users. The state is additionally asking the court to bar OpenAI from overstating the safety of ChatGPT and its teen version, and to require the company to disclose known risks and provide effective parental notifications.

The central allegation is that the chatbot is built around continuous, human-like conversation. The lawsuit describes this style as a combination of anthropomorphism and sycophancy: the system allegedly tends to affirm users in ways that keep them engaged, even when a teenager needs an adult’s intervention rather than more messages on the screen. The state identifies conversations about suicide, psychosis, self-harm and violence as especially dangerous. These are allegations by the state that OpenAI has yet to contest in court.

ChatGPT for Teens—a version of the service for users ages 13 to 18, launched in August 2026—features prominently in the complaint. Oregon alleges that linking a parent’s account to a teen’s account remains optional, meaning the advertised parental features may simply never be activated. The complaint also cites testing by Common Sense Media: according to the organization, parents received no alerts during conversations showing signs of crisis, and in some cases the bot did not suggest contacting a helpline or a professional.

According to the attorney general’s office, the problem starts even earlier—with the user’s age. OpenAI’s rules allow teenagers aged 13 and older to use ChatGPT with a parent or guardian’s consent, but the lawsuit says the company did not previously verify that consent during registration. The state also questions the effectiveness of a later age-prediction system, which is supposed to identify minors based on account and behavioral patterns. For a digital service where age may amount to a birth date entered by the child, this is not a technical detail but the foundation of the entire protection model.

Rayfield does not link the lawsuit to any confirmed incident in Oregon itself. In an interview with Oregon Public Broadcasting, he said the decision followed reports of incidents across the country and concern that technology is advancing faster than consumer protections. The lawsuit is part of a new attorney general’s initiative, Keeping Kids Safe, which brings together the state’s online child-safety cases, investigations into other platforms and an expansion of units combating the exploitation of minors.

A representative for OpenAI said the company intends to work with experts on additional safeguards, including parental controls, break reminders and age-based model settings, and supports rules that apply across the industry rather than to a single company. But Oregon’s lawsuit raises a more uncomfortable question: is it enough to fix individual failures if the chatbot is increasingly becoming a child’s conversation partner, adviser and invisible companion during hours when no adult is nearby? The answer will now be sought not only by engineers, but by a court.

Based on: Oregon sues OpenAI for $2 billion, says ChatGPT puts kids in harm’s way